A certain number of techniques are known, which enable tracking of content use on a terminal, among which are included the processing of content server logs, techniques using markers, techniques implementing scripts or else techniques implementing application extensions in the terminal.
1. Processing of Server Logs
This technique consists in applying a computer programme to the content server logs, which generally uses the HTTP protocol. This programme adds a line of textual information to the end of the log file upon each content request.
This line contains a set of information such as the origin of the request, the protocol and the type of request, the address of the requested content, success or an error code, the type of browser, the original content of the request . . . .
By cross-referencing between the various lines added, and by assuming that there is a single user per origin of the request, the downloading information for a content family by a user can be reconstructed.
The disadvantages of this technique of the prior art lies in the fact that it is not possible to differentiate between content downloaded onto a terminal, i.e., the request of which was followed by success, and used by the terminal, and content pre-loaded onto the terminal with a view to later consumption or use, when the terminal is in a mode unconnected to the communication network or offline.
Another disadvantage is that it is not possible to track content consumed multiple times, when the terminal is in disconnected mode.
Neither does this technique enable differentiation between the downloading of content and content use activity, such as partial reproduction of the content, backing up certain elements of the content . . . .
2. Techniques Implementing Markers or “Tags”
The principle of these techniques (“Accrue Insight,” “ESTAT'PRO,” “WEBandSTATS,” “Weboscope,” “Webtrends Reporting Series,” “Wysistats” . . . ) is to call up content, an image or a URL, using parameters enabling the user to be “tracked,” and to pull up available information via JavaScript.
They therefore require code to be inserted into the access pages of the content server.
They thus enable access to a service to be traced, even if it is carried out entirely in a description language with script, such as Flash.
For example, in content anticipating user interactivity, which enables reading of an element of the scene to be initiated, a message can also be sent to an external server when the action is initiated, in order to record this reading.
However, one disadvantage of these techniques is that they require an outside server in order to record the tracing request.
In addition, these techniques do not work in connected mode.
3. Techniques Implementing Scripts
These techniques implement an interpreted programme, which is written in a language such as “ECMA-Script” or “Action-Script,” which creates the tracking information and sends same to the server responsible for processing this information.
One disadvantage of these techniques lies in the fact that this information cannot be stored on the terminal, for security reasons.
As a matter of fact, it is unadvisable to perform “scripting” and at the same time write into the memory of a terminal. For example, security restrictions are specified for the use of Java applets when they come from the Web (no permanent file writing on the terminal).
These techniques do not work, therefore, when the terminal is in disconnected mode.
4. Reader Application Extensions
These techniques consist in installing a special reader on the terminal, e.g., such as the iPod, which has a proprietary Apple reader, and which, for example, stores the information relating to the number of musical content consumptions, and then synchronizes same with the computer.
One disadvantage of this technique lies in the use of a proprietary reader, which provides the programme enabling tracking.